ABSTRACT

With the growth of using the internet and the large amount of actual-time data created and shared on social media sites and applications have raised the risk of spreading harmful or misunderstanding content, specifically on Facebook. Fake news, or intentionally misleading information, has become a significant problem can have serious consequences, including harm to individuals and damage to society.

To address this issue, we propose a system for detecting fake news in Kurdish language social media using a combination of natural language processing (NLP) techniques and machine learning algorithms. We created a dataset of Kurdish language news articles, both real and fake. The dataset will include a variety of article types, including political, social, and entertainment news. Also, we created a framework Research procedure to make sure from the data we have created, the system will be trained on and also, the system will be tested on a separate dataset to evaluate its performance in detecting fake news and classifying articles into the appropriate categories.

Results of the proposed study demonstrated that Passive-Aggressive Classifier (PAC) outperformed the other classifiers on both datasets with an accuracy score of 93.0 percent and other classifiers were less in some precents that show high accuracy as well since it is 90.0 percent. Also, the results of this research will contribute to the development of technology that can help combat the spread of fake news on social media and promote the dissemination of accurate information. 

Keywords: Fake News Detection, Kurdish Language, Machine learning, Classifiers, Passive-Aggressive.